A televised programmed instruction series for teaching Spanish to fifth and sixth grade students in Denver, Colorado, was evaluated. Three instructional methods were studied: programmed instruction with a proctor; conventional instruction with a trained teacher; and instructional program plus a trained teacher. The following factors were explored, in addition to student achievement: (1) overall differences among the instructional groups; (2) changes that took place among the classes and what kinds of classes profited from each type of instruction; (3) relationships among student characteristics in each instructional group; (4) how teacher attitudes toward the subject matter and instructional techniques were affected by their experiences; and (5) the relationship between teacher and class characteristics. It was concluded that the instructional program was effective, especially with a trained teacher. Both Spanish pretest and IQ were valid predictors of achievement. Teachers' experience with programmed instruction was associated with the desire to use the program. Students with high ability who were interested in independent study tended to be taught by the program plus a teacher. Presented at 48th Annual Convention, NAASP.